The accumulation of the beta-amyloid peptide (Abeta) is a central event in the pathogenesis of Alzheimer's disease (AD). Abeta removal from the brain by immune therapy shows promising potential for the treatment of patients with AD, although the mechanisms of the antibody action are incompletely understood. In this study we compared the biological activities of antibodies raised against various Abeta fragments for Abeta reduction in vitro and in vivo. Antibodies against Abeta enhanced the uptake of Abeta42 aggregates up to 6-fold by primary microglial cells in vitro. The kinetics of Abeta42 uptake varied considerably among antibodies. Based on the activity to mediate Abeta42 uptake by microglial cells, we identified a bioactive antibody that significantly reduced Abeta42 levels in the brains of transgenic mice with neuronal expression of an AD-related mutated amyloid precursor protein. This effect depended on the epitopes recognized by the antibody. Our data suggest that the ability to facilitate Abeta42 uptake by primary microglia cells in vitro can be used to predict the biological activity of the antibody by passive immunization in vivo. This protocol may prove useful for the rapid validation of the activity of antibodies designed to be used in immune therapy of AD.